The case for great smart home management has already been made. Consumers love it, they’re willing to pay extra for it, and it can also dramatically reduce operator costs. But with an increasingly crowded field of vendors, how can ISPs be sure they’re making the right procurement choices?

In this series of blog posts, I outline five key questions operators should ask themselves to avoid making an expensive mistake:

Exactly how smart is the smart home security?

Endpoint security is essential to countering external attacks, but the firewalls in your existing routers aren’t up to the job. Many will only review inbound traffic, offering no protection to data leaving the home network. They may also require configuration and maintenance far beyond the average consumer’s ability.

Most smart home management systems use blacklists to block traffic from known malicious sites, however you also need to address internal and outgoing traffic. Will a system prevent connected devices from initiating contact with blacklisted sites, or being used to attack other in-home devices? And can it perform real-time blocking at both IP address and DNS level?

To get the most from your existing investments, consider whether each smart home system under consideration can automatically ensure subscribers’ routers have the latest firmware, firewall settings, and port usage restrictions.

Tackling the unknown threats

No matter how great your threat database, attacks from unknown sources are a huge risk. That’s why smart homes need robust anomaly detection – the ability to block or quarantine devices that begin behaving outside their normal parameters. But this presents another challenge: how to identify what constitutes “normal” behavior so you know when the device is acting strangely?

With thousands of different devices on sale, and more being added daily, no vendor can manually maintain a database of acceptable traffic patterns for every device.

Machine learning brings better blocking

True smart home management requires accurate IoT fingerprinting to correctly identify all the devices connected in a home. Once you’ve got that clear picture, Artificial Intelligence is useful for rapidly profiling each device – regardless of manufacturer or model – and establish a broad picture of “normal” traffic patterns. This data enables swift identification and blocking of any activity that deviates from the norm.

For each solution under consideration, check how accurately it differentiates a CCTV camera from a smart fridge. Then confirm if it can automatically block the camera from contacting inappropriate websites or stop the fridge from communicating with other devices in the home.

Finally, with up to 79% of consumers admitting to fears about data security in their smart home, it’s important to know whether any system you evaluate can quarantine devices to prevent personal data from leaving the network in the event of a breach. And consider privacy implications carefully. Will your subscribers like the content of their emails being examined by Deep Packet Inspection techniques?

Four more key evaluation questions for smart home management systems

You can read more on the other topics to consider in your smart home security procurement process in my upcoming blogs. They range from IoT Fingerprinting and hardware considerations to open standards and feature set. And there’s more information here on how Irdeto’s smart home management solution, Trusted Home, meets these challenges.

Bengt Jonsson | SVP Sales